Benjamin D Edwards

US Patent:

20210150042, May 20, 2021

Filed:

Nov 15, 2019

Appl. No.:

16/685474

Inventors:

- Armonk NY, US
Frederico Araujo - White Plains NY, US
Teryl Taylor - Danbury CT, US
Marc Phillipe Stoecklin - White Plains NY, US
Benjamin James Edwards - Ossining NY, US
Ian Michael Molloy - Westchester NY, US

Assignee:

International Business Machines Corporation - Armonk NY

International Classification:

G06F 21/62
H04L 9/08
G06F 17/16
G06N 20/00

Abstract:

A neural network is trained using a training data set, resulting in a set of model weights, namely, a matrix X, corresponding to the trained network. The set of model weights is then modified to produce a locked matrix X′, which is generated by applying a key. In one embodiment, the key is a binary matrix {0, 1} that zeros (masks) out certain neurons in the network, thereby protecting the network. In another embodiment, the key comprises a matrix of sign values {−1, +1}. In yet another embodiment, the key comprises a set of real values. Preferably, the key is derived by applying a key derivation function to a secret value. The key is symmetric, such that the key used to protect the model weight matrix X (to generate the locked matrix) is also used to recover that matrix, and thus enable access to the model as it was trained.

US Patent:

20200050945, Feb 13, 2020

Filed:

Aug 7, 2018

Appl. No.:

16/057706

Inventors:

- Armonk NY, US
Wilka Carvalho - Los Angeles CA, US
Heiko H. Ludwig - San Francisco CA, US
Ian Michael Molloy - Chappaqua NY, US
Taesung Lee - Ridgefield CT, US
Jialong Zhang - White Plains NY, US
Benjamin J. Edwards - Ossining NY, US

International Classification:

G06N 3/08
G06N 3/04

Abstract:

One embodiment provides a method comprising receiving a training set comprising a plurality of data points, where a neural network is trained as a classifier based on the training set. The method further comprises, for each data point of the training set, classifying the data point with one of a plurality of classification labels using the trained neural network, and recording neuronal activations of a portion of the trained neural network in response to the data point. The method further comprises, for each classification label that a portion of the training set has been classified with, clustering a portion of all recorded neuronal activations that are in response to the portion of the training set, and detecting one or more poisonous data points in the portion of the training set based on the clustering.

US Patent:

20170295970, Oct 19, 2017

Filed:

Apr 14, 2017

Appl. No.:

15/487984

Inventors:

Benjamin Edwards - Washington DC, US

International Classification:

A47G 21/00
A47G 21/00

Abstract:

An improved temperature regulating dinnerware set is a dinnerware set with each dinnerware piece having a heating element and a control member for controlling the temperature of the top surface of the piece. Each piece is adapted to connect to a power source to control the embedded heating element. An attachable cover is provided for each piece of the dinnerware set to help maintain the temperature.